A T-shirt sniffing test has revealed that women unwittingly prefer the smell of men who have similar genes to their dads. But this is no Freudian Oedipal complex.

Instead, it appears to be a tactic in a poorly understood evolutionary game, where the prize is either greater resistance to disease, or an unconscious ability to spot distant relatives in a sea of strangers.

The genes in question form part of the major histocompatibility complex, or MHC, and encode various components of the immune system. These genes are thought to be tightly linked to others that dictate our natural odour.

Research on animals has shown that female mice sniff out males with different MHCs to their own, prefering them to mates with a similar genetic make up. Women were also thought to do the same, according to one study in which women sniffed T-shirts worn for a couple of nights by men.
The same, but different

Now a new study paints a more complicated picture. Martha McClintock, Carole Ober and a team at the University of Chicago studied 49 women whose MHC genes and parents' MHC genes were known. As in the earlier T-shirt study, the women sniffed T-shirt odours, but this time they had no idea what they were smelling. They were asked to say which odours they would prefer if they had to smell them all the time.

Surprisingly, the women preferred the odours of men who shared the same type of a few MHC genes, or alleles, with themselves. The most appealing odour donors shared 1.4 alleles on average, whereas the least appealing shared 0.6 alleles. What's more, these matching alleles were ones the women had inherited from their fathers and not from their mothers.

That goes against the prevailing theory that outbreeding is always best. Going for a mate with different immune system genes to your own should ensure that your children have the widest possible arsenal with which to attack pathogens. Also, the rarer their MHC, the less likely it is that evolving pathogens will be able to outsmart them.
Limited inbreeding

But McClintock thinks that interpretation is too narrow. Limited inbreeding can work, as it may actually make sense to stick with combinations of genes that are known to successfully fight disease. "There's an intermediate number of matches that's probably optimal," she says.

Wayne Potts of the University of Utah in Salt Lake City has a different explanation. Although mice prefer mates with different MHC genes, they go for nest mates with a similar genetic make-up, probably to ensure they are near their kin. Women may be attracted to their father's odours for a similar reason - reflecting an ability to home in on relatives using smell.

For instance, he says that Ober's own studies show that women tend to marry MHC-dissimilar men (New Scientist, 10 February 2001, p 36). "It is probably more reliable to draw conclusions ... from marriage patterns," he comments, "than from odour preference tests where boxes with odiferous, unknown contents are briefly sniffed."